Objective: Approximately 70% of the time spent during the construction of a modular building is devoted to the onsite assembly of the prefabricated components. Because of the limited accuracy of traditional crane motion control, insufficient rigidity of the boom and rope, and susceptibility of cranes to outdoor environments, accurate placement of components in the target area using a crane alone is difficult. Repeated positioning of a component with the assistance of multiple workers is required before installing the component. Repeated steps of lifting and lowering components rely on a considerable amount of labor and affect construction efficiency. To solve the problems of traditional installation methods that rely excessively on manual labor for the positional adjustment of lifted components, a robot-assisted component installation system is proposed in this study. Methods: Construction sites are far more complex than structured factory manufacturing environments; therefore, construction robots have distinct technical characteristics compared with industrial robots. A robot-assisted installation method was designed based on an analysis of the technical characteristics of construction robots. After the initial alignment of a component using a crane, the two robots cooperate to adjust the position and orientation of the component accurately. Thus, automatic installation of the components can be realized. The procedure for conducting the entire installation-assisted task of the robot was illustrated in a pseudocode form for a series of actions, including positioning the robots and the reserved hole, threading the hole, and pushing components. To reduce the cycle time and costs for the development of such a robot, a prototype model of a robot-assisted component installation system was created on a computer and virtual prototyping was used to simulate and analyze the kinematic and dynamic properties of the model within a virtual system with real environmental properties. Moreover, the calculation results of the virtual prototype provided important data support for equipment selection and component design in subsequent test sessions. Results: The motion state and the change of contact force between the tool mounted on the end flange and the prefabricated component of the two robots during the task execution were accurately modeled. Based on the simulation results, the feasibility of the robot replacing workers to complete the positional adjustment of components and the effective reduction of the end load of the robot was verified. With the robot-assisted component installation system built in the laboratory, which was identical to the virtual prototype model, the test of installing a precast panel in a predetermined area was conducted several times. The installation position deviations observed in all tests were less than the limit values of the quality acceptance standards in China. Conclusions: The actions of the two robots for assisting in the accurate installation of components can be smoothly realized. The simulation calculations and experimental results demonstrate the advancement of the proposed robot-assisted component installation system for improving component installation accuracy and saving labor. Because of severe labor shortages and sharp increases in labor costs, the proposed method provides economic benefits that become more evident as the number of component installation tasks increases. Research on installation-assisted robots has substantial reference value and application prospects for recent research progress on construction robots and practical engineering problems. With the future development of intelligent control systems, the automation level of the proposed method can be considerably improved to realize truly unmanned construction.